Receptors for toxic shock syndrome toxin-1 and staphylococcal enterotoxin A on human blood monocytes

Can. J. Microbiol. Downloaded from www.nrcresearchpress.com by NEW YORK UNIVERSITY on 11/12/14 For personal use only.

RAYMOND H. SEE Division of Infectious Diseases, Departments of Medicine and Pathology, University of British Columbia, Vancouver, B.C., Canada V5Z 3J5 and Canadian Bacterial Diseases Net work, Vancouver, B. C., Canada Department of Pathology, University of British Columbia, Vancouver, B.C., Canada V6T I W5 and Terry Fox Laboratory, British Columbia Cancer Research Center, Vancouver, B. C., Canada AND

ANTHONY W.

CHOW

Division of Infectious Diseases, Departments of Medicine, Microbiology, and Pathology, University of British Columbia, Vancouver, B. C., Canada V6T I W5 and Canadian Bacterial Diseases Network, Vancouver, B.C., Canada Received August 16, 1991 Revision received February 18, 1992 Accepted March 6, 1992 SEE, R.H., KRYSTAL, G., and CHOW,A. W. 1992. Receptors for toxic shock syndrome toxin-1 and staphylococcal enterotoxin A on human blood monocytes. Can. J . Microbiol. 38: 937-944. Staphylococcal toxic shock syndrome toxin-1 (TSST-1) as well as staphylococcal enterotoxin A (SEA) and B (SEB) have recently been shown to bind directly to the class I1 major histocompatibility antigen, HLA-DR. Whereas others have characterized TSST-1 and SEA binding to HLA-DR on transfected L cells or B lymphoma cell lines, we sought evidence for direct binding of TSST-1 and SEA to HLA-DR on purified human monocytes. A single class of highaffinity receptors was found for both TSST-1 (dissociation constant ( K , ) 40 nM, 3.4 x lo4 receptors per cell) and toxins SEA (K, 12 nM, 3.2 x lo4 receptors per cell) on normal human monocytes. Affinity cross-linking of 125~-labeled to monocytes revealed the presence of two membrane protein subunits with molecular masses consistent with the a and p chains of human HLA-DR (35 and 28 kDa, respectively). The anti-HLA-DR monoclonal antibody L243, but not L203 or 2.06, inhibited radiolabeled toxin binding to human monocytes and neutralized the mitogenic response of human T lymphocytes to both toxins. However, L243 was consistently more effective in blocking radiolabeled TSST-1 than SEA binding to human monocytes from the same donors, suggesting that TSST-1 and SEA may be binding to overlapping epitopes rather than to the same epitope on HLA-DR. Because TSST-1 and SEB bind to distinct epitopes on HLA-DR and because SEA cross competes with both TSST-1 and SEB on the HLA-DR receptor, we postulate that SEA occupies a binding site within HLA-DR that overlaps both TSST-1 and SEB. Future studies focused on receptormediated binding of these toxins to human monocytes and T lymphocytes from normal donors and toxic shock syndrome patients may reveal the underlying anomalies that predispose particular individuals to toxic shock syndrome. Key words: monocytes, staphylococcal toxic shock syndrome toxin-1, receptors, HLA-DR, staphylococcal enterotoxin A. SEE, R.H., KRYSTAL, G., et CHOW,A. W. 1992. Receptors for toxic shock syndrome toxin-1 and staphylococcal enterotoxin A on human blood monocytes. Can. J. Microbiol. 38 : 937-944. I1 a Cte demontre rkcemment que la toxine du syndrome de choc toxique du staphylocoque (TSST-1) ainsi que les enterotoxines SEA et SEB se lient directement a l'antigene majeur d'histocompatibilite de classe 11, HLA-DR. Poursuivant les travaux d'autres chercheurs qui ont caracterise la liaison de TSST-1 et de SEA a HLA-DR sur des cellules L transfectees ou sur des lignees cellulaires de lymphome a cellules B, les auteurs ont recherche l'evidence d'une liaison directe de TSST-1 et de SEA a HLA-DR sur des monocytes humains purifies. Une seule classe de recepteurs a grande affinite pour TSST-1 (Kd 40 nM, 3,4 x lo4 recepteurs par cellule) et SEA (K, 12 nM, 3,2 x lo4 recepteurs par cellule) a ete retrouvee sur des monocytes humains normaux. L'affinite de liaison sur les monocytes, des toxines marquees a l'iode ( ' 2 5 ~ )a, revel6 la presence de deux sous-unites de proteines membranaires possedant des masses moleculaires compatibles avec les chaines a et /3 de HLA-DR humain (respectivement 35 et 28 kDa). L'anticorps monoclonal anti-HLA-DR L243, contrairement aux anticorps L203 ou 2.06, a inhibe la liaison des toxines radio-marquees aux monocytes humains, et a neutralise la reponse mitogenique des lymphocytes T humains aux deux toxines. Cependant, L243 a ete regulierement plus efficace pour bloquer la liaison de TSST-1 radio-marquee aux monocytes humains provenant des memes donneurs, que pour SEA; ce resultat suggere que TSST-1 et SEA pourraient se lier a des epitopes en chevauchement plut8t qu'au meme epitope sur HLA-DR. I1 a ete demontre que TSST-1 et SEB se lient a des epitopes distincts ' ~ u t h o rto whom all correspondence should be addressed at the Division of Infectious Diseases, G. F. Strong Research Laboratories, Vancouver General Hospital, 2733 Heather Street, Vancouver, B.C., Canada V5Z 355. Printed in Canada / Imprime au Canada

CAN. J. MICROBIOL. VOL. 38, 1992

Can. J. Microbiol. Downloaded from www.nrcresearchpress.com by NEW YORK UNIVERSITY on 11/12/14 For personal use only.

sur HLA-DR et que SEA competitionne a la fois avec TSST-1 et SEB sur le recepteur de HLA-DR; par consequent, les auteurs postulent que SEA occupe un site de liaison qui chevauche ceux de TSST-1 et de SEB a l'interieur de HLA-DR. Des etudes ulterieures, dirigees sur la liaison determinee par le recepteur de ces toxines aux monocytes humains et aux lymphocytes T obtenus a partir de donneurs normaux et de patients victimes du syndrome de choc toxique, peuvent reveler les anomalies sous-jacentes qui predisposent des individus particuliers au syndrome de choc toxique. Mots elks : monocytes, recepteurs, HLA-DR, enterotoxine A du staphylocoque, toxine du syndrome de choc toxique du staphylocoque. [Traduit par la redaction]

could not be easily resolved but could, in part, be attributed Toxic shock syndrome (TSS) is a multisystem disease to methodologic issues, including purity of toxin preparaassociated with Staphylococcus aureus infection. A 22-kDa tions and method of isolation or purity of monocyte or T cell exoprotein, TSS toxin- 1 (TSST- I), has been strongly implifractions from human PBMCs. For example, we found that cated as the primary agent in the pathogenesis of TSS. In the induction of TNFa and IL-1fl by highly purified TSST-1 addition to TSST- 1, staphylococcal enterotoxins (especially required the presence of both human monocytes and T cells, SEA and SEB) may also be involved in some TSS cases whereas commercial sources of TSST-1 containing impurities (Schlievert 1986; Whiting et al. 1989). We previously dem(Toxin Technology, Madison, Wis.) induced TNFa from onstrated a significantly higher seroconversion rate to SEA human monocytes in the absence of T cells (Kum et al. 1992; as well as TSST-1 by using paired acute and convalescent See et al. 1992b). In the present study, we report our findsera from TSS patients compared with non-TSS-associated ings of TSST-1 and SEA binding to highly purified normal control patients with S. aureus infection (Whiting et al. human monocytes fractionated from PBMCs by density 1989). We further observed that a single clone of S. aureus, centrifugation over Percoll. Specific binding of both toxins which produces both TSST-1 and SEA, is associated with to human monocytes was demonstrated in competitive bindmost urogenital TSS cases (71%) in contrast to nonurogenital TSS (24%) or non-TSS-associated control ing studies, and by cross-linking experiments with or without patients (12%) (Chang et al. 1991). Both TSST-1 and SEA prior treatment of monocytes with excess unlabeled homologous toxin, or monoclonal antibodies to HLA-DR, the have potent mitogenic effects on human T lymphocytes and stimulate the release of lymphokines such a ~ - ~ - i n t e r f e r o n putative monocyte receptor for these toxins. The nature of TSST-1 and SEA binding domains to human monocyte (Jupin et al. 1988; Poindexter and Schlievert 1985). Both receptors was also examined by inhibition of toxin binding, toxjns also activate monocytes to release monokines such using a panel of three monoclonal antibodies to HLA-DR. as interleukin-1 (IL-1) and tumour necrosis factor (TNF) Finally, the effect of inhibition of toxin binding to human (Ikejima et al. 1984; Jupin et al. 1988; Parsonnet et al. PBMCs on TSST-1 and SEA-induced T cell proliferation 1986). The induction of these biologic response modifiers was studied using anti-HLA-DR monoclonal antibodies. may explain some of the characteristic features observed These results clearly demonstrate the presence of specific among TSS patients, such as fever, hypotension, and shock. receptors for both TSST-1 and SEA on normal human The mechanism by which TSST-1 and staphylococcal enterotoxins activate monocytes is poorly understood. monocytes, and that the binding domains of these toxins TSST-1, SEA, and SEB have been shown to bind to spefor human monocytes have overlapping epitopes in the cific receptors on human peripheral blood mononuclear cells HLA-DR receptor. (PBMCs) (Mourad et al. 1989; Poindexter and Schlievert 1987; Scholl et al. 1989a; See et al. 1990). We have recently Materials and methods shown that TSST-1 and SEA bind to separate or overlapTSST-I and SEA ping epitopes on the same receptor on human PBMCs (See TSST-1 was purified from culture supernatants of S. aureus et al. 1990). The receptors for these toxins have now been MN8 by ion-exchange chromatography, chromatofocusing, and identified as the class I1 major histocompatibility complex gel filtration, as previously described (Rosten et al. 1989; Kum et al. (MHC) antigen, using Epstein-Barr virus (EBV) trans1992). Purity of TSST-1 was > 95%, as determined by silver stainformed B lymphoma (Raji) cells (Fischer et al. 1989), class I1 ing of sodium dodecyl sulfate (SDS) - polyacrylamide gels (Laemmli 1970) and by immunoblotting (Burnette 1981) with rabMHC-bearing T cell lines (Scholl et al. 1989a), and class I1 bit polyclonal antisera raised against crude culture filtrate of MN8 MHC-transfected L cells (Scholl et al. 1989b; Uchiyama and with pooled normal human serum. Highly purified (>95%) et al. 1989). TSST-1 receptors have also been found on staphylococcal enterotoxin A (SEA) was purchased from Toxin human B lymphocytes (Mourad et al. 1989; Scholl et al. Technology, Inc. (Madison, Wis.). TSST-1 and SEA were iodinated 1989a; Uchiyama et al. 1989). Binding of TSST- 1 and SEA by the chloramine T method as previously described (See et al. to purified human monocytes, however, has not been vig1990). Specific activity of the radioiodinated toxins ranged from orously investigated. Poindexter and Schlievert (1987) 30 to 55 pCi (1 Ci = 37 GBq) per microgram of protein, with examined 125~-labeled TSST-1 binding to T and B lympho< 10% loss of mitogenic activity as determined by their effect on cytes and monocytes fractionated from normal human human PBMCs. Analysis of 12'1-labeled TSST- 1 and '*'I-labeled PBMCs. They reported that TSST-1 bound specifically to SEA by SDS - polyacrylamide gel electrophoresis (SDS-PAGE) and autoradiography confirmed their apparent molecular mass as human T cells but were unable to demonstrate specific bind22 and 28 kDa, respectively. ing of TSST-1 to human B cells or monocytes. Scholl et al. (1989a), on the other hand, reported specific binding of Monoclonal antibodies TSST-1 to human monocytes but not resting T cells. Monoclonal antibodies against monomorphic determinants of Uchiyama et al. (1989) also could not detect TSST-1 bindHLA-DR (L243, L203, and 2.06) as well as the monocyte-specific ing to human T cells, while Mollick et a/. (1989) were unable CDl l b antigen (OKMI) were produced by hybridomas obtained from the American Type Culture Collection, Bethesda, Md. The to detect SEA binding to human T cells. These differences

939

SEE ET AL.

Can. J. Microbiol. Downloaded from www.nrcresearchpress.com by NEW YORK UNIVERSITY on 11/12/14 For personal use only.

antibodies were purified from ascitic fluid by ammonium sulfate precipitation followed by protein G chromatography, using the MAb Trap kit (Pharmacia Fine Chemicals, Dorval, Que., Canada). Purification of human monocytes Fresh human PBMCs were obtained by centrifugation of plateletpheresis buffy coats from healthy adult donors over Histopaque 1.077 (Sigma Chemical Co., St. Louis, Mo.). Cells at the interface were washed five times in Hank's balanced salt solution and additionally separated into T cell and non-T-cell subpopulations by rosetting with sheep red blood cells as described (Madsen and Johnson 1979). Monocytes were then separated from B lymphocytes by density centrifugation over Percoll as outlined by de Boer et al. (1981). In brief, the non-T-cell fraction was resuspended in RPMI 1640 with 1070 heat-inactivated fetal calf serum (FCS) at a concentration of 5 x lo7 cells/mL. The cells were then mixed with Percoll (Pharmacia Fine Chemicals, Dorval, Que., Canada) to give a final specific gravity of 1.062 g/mL. One millilitre of RPMI 1640 with 1070 FCS was gently layered on top of each suspension. The gradient was then centrifuged at 850 x g for 15 min. The monocyte-containing interface was removed, washed three times, and suspended in RPMI 1640 containing interface was removed, washed three times, and suspended in RPMI 1640 containing 1% bovine serum albumin (binding buffer). Purity of the monocyte preparation was 2 90% as assessed by nonspecific esterase staining of preparations (Yam et al. 1971). Trypan blue exclusion showed > 95% viability of cells. Binding assay Binding assays were performed by suspending human monocytes at 5 x lo7 to 1 x lo8 cells/mL. Approximately lo7 monocytes were incubated at 4°C with various concentrations of ','I-labeled TSST-1 or ','I-labeled SEA in final volumes of 200 pL. For binding inhibition assays, monocytes were preincubated for 60 min at 4°C with anti-HLA-DR monoclonal antibodies (L243, L203, 2.06) or a monocyte-specific isotype control monoclonal antibody (OKMl) before the addition of the radiolabeled toxin. The cell suspensions were gently rotated for at least 1 h, and unbound radiolabeled toxin was then separated from the cells by transferring triplicate 60-pL portions of assay mixture to precooled microcentrifuge tubes containing 280 pL of a mixture of dibutyl phthalate and dioctyl phthalate oils (1.1:l) (BDH, Vancouver, B.C.). The tubes were centrifuged at 16 000 x g for 1 min and then immediately frozen at - 70°C, and the tube tips containing the cell pellets were excised. Cell-associated radioactivity was determined with a Searle 1185 gamma counter. Nonspecific binding was determined by adding a 100-fold or greater molar excess of unlabeled toxin. All data were corrected for nonspecific binding. Binding data were analyzed using the LIGAND program (Munson and Rodbard 1980). Receptor cross-linking Cross-linking studies were performed using a modification of the method previously described (Sorensen et al. 1986). Specifically 1 x lo7 human monocytes were incubated with 60 nM of either labeled TSST-1 or ','I-labeled SEA for at least 1 h at 4°C. The cells were washed twice and resuspended in 200 pL of ice-cold phosphate-buffered saline, pH 7.4. Disuccinimidyl suberate (DSS) (Pierce, Rockford, Ill.), freshly dissolved in dimethyl sulfoxide at 40 mM, was added to give a final concentration of 1 mM. After rotating at 4°C for 30 min, the reaction was stopped by the addition of 200 pL of 50 mM Tris-HC1, pH 7.4, and the cells were pelleted at 250 x g for 5 min. Monocytes were then resuspended in 1 mL ice-cold 10 mM Tris-HC1, pH 7.4, containing 10 pg leupeptin/mL, 100 kallikrien inhibitor units aprotonin/mL, and 2 mM freshly added phenylmethylsulfonyl fluoride (PMSF). After swelling for 10 min at 4"C, cells were lysed by passing through a 26-gauge needle six times. Released nuclei and unbroken cells were removed by centrifugation at 800 x g for 5 min at 4°C. Membranes were then pelleted at 16 000 x g for

0

0.2

0.4

0.6

0.8

1

1.2

1.4

Bound TSST-1 (nM)

FIG. 1. Representative Scatchard plot of ','I-labeled TSST-1 binding to purified human monocytes (lo7 cells). Cells were incubated with 2-270 nM radiolabeled toxin and processed as described in the text.

30 min. The resulting pellet was dissolved in 40 pL of SDS - sample buffer containing 2% SDS, 5% 2-mercaptoethanol, and 10% glycerol, boiled for 3 min, and analysed by SDS-PAGE using a 12% separating gel. After electrophoresis, the gel was dried and exposed to Kodak XOMAT-AR film at -70°C in the presence of Cronex lighting plus intensifying screens. For monoclonal antibody inhibition experiments, monocytes were pretreated with monoclonal antibodies for 1 h at 4"C, washed, and cross-linked to radiolabeled toxin as described above. Mitogenicity assay Mitogenicity assays were performed as previously described (Poindexter and Schlievert 1985). Briefly, human PBMCs were suspended at a concentration of 3 x lo6 cells/mL in complete RPMI 1640 supplemented with 1070 heat-inactivated fetal calf serum, 25 mM Hepes, 2 mM L-glutamine, and 20 pg polymyxin B sulfate/mL. Cells (3 x 10') were cultured in 200-pL volumes with various concentrations of TSST-1 and SEA in round-bottom 96-well tissue culture plates (Falcon Plastics) for 3 days at 37°C in 5% CO,. For monoclonal antibody inhibition studies, PBMCs were preincubated with the indicated concentrations of anti-HLADR monoclonal antibodies or control antibody OKMl for 1 h at 37°C before toxin stimulation. At 48 h, cells were pulsed with 1 pCi of [3~]thymidine (5 Ci/mmol; Amersham, Oakville, Ont.) and harvested 18 h later onto glass-fiber filter paper, using an automatic harvester (Skatron, Norway). Samples were counted in a liquid scintillation counter (Beckman LS 1800). All assays were carried out in quadruplicate, and controls included PBMCs alone, PBMCs plus monoclonal antibody, and PBMCs plus toxin.

Results We and others have previously demonstrated the presence of TSST-1-specific receptors on human PBMCs (Mourad et al. 1989; Scholl et al. 1989a; See et al. 1990; Uchiyama et al. 1989). In the present study, we demonstrate that purified human monocytes have specific receptors for TSST- 1 and the related staphylococcal enterotoxin A (SEA). Optimal conditions (buffers, equilibrium period, etc.) for binding of 12'1-labeled toxins were used as outlined previously (See et al. 1990). Figure 1 shows a representative Scatchard plot for TSST-1 as calculated by the LIGAND program (Munson and Rodbard 1980). Analysis of 1 2 5 ~ labeled TSST-1 binding to monocytes from five different donors revealed the presence of 34 000 k 8658 TSST-1 receptors per cell, with a dissociation constant (Kd) of

940

CAN. J. MICROBIOL. VOL. 38, 1992

TABLE1. Receptor numbers and dissociation constants for TSST-1 and SEA on normal human blood monocytes TSST- 1

Can. J. Microbiol. Downloaded from www.nrcresearchpress.com by NEW YORK UNIVERSITY on 11/12/14 For personal use only.

Donor No.

100

SEA

Kdb

Ra

Kd

R

number o f receptors per cell, as determined by the LIGAND program. constant (nanomolar), as determined by the L I G A N D program. 'nd, not determined. d ~ e a n+ standard deviation for five donors. O R ,

-f3

0OKM 1 L243 L203 2.06 FIG. 2. Inhibition of (A) 125~-labeled TSST-1 or (B) '25~-labeled SEA binding to human monocytes by anti-HLA-DR monoclonal antibodies. Monocytes were preincubated with either antiHLA-DR monoclonal antibodies (L243, L203,2.06) or a monocytespecific isotype control monoclonal antibody (OKM1) before the addition of 30 nM '25~-labeledTSST-1 or '25~-labeledSEA. The percentage of radiolabeled toxin bound in the absence of antibodies was normalized to 100% (ordinate). Values represent the mean of triplicate determinations k the standard deviation. Nonspecific counts have been subtracted. MAb, monoclonal antibody.

b ~ d dissociation ,

30 -

a

D

I

I

TSST - I

I

1

SEA

FIG. 3. Characterization of TSST-1 and SEA receptors on human monocytes by affinity cross-lining. Monocytes were incubated with 60 nM of 125~-labeled TSST-1 at 4OC, treated with DSS, and analysed by SDS-PAGE under nonreducing conditions. 40 k 15 nM (Table 1). Similar receptor numbers and disCells were exposed to 12'1-labeled TSST-1 alone (lane 1) or sociation constants were found for SEA (Table 1). For both 100-fold molar excess of unlabeled TSST-1 plus '25~-labeled toxins, the LIGAND program indicated only one class of TSST-1 (lane 2). Similarly, '25~-labeledSEA was cross-linked to receptors. To determine the identity of the receptors, antimonocytes (lane 5) in the presence of excess unlabeled SEA (lane 6), HLA-DR monoclonal antibodies were preincubated with Monocytes were preincubated with 100 pg L243/mL (lanes 3 and 7) or OKMl (lanes 4 and 8) before cross-linking with 125~-labeled monocytes before the addition of radiolabeled toxin. Results TSST-1 or '25~-labeledSEA. Arrowheads A and B represent the show that the anti-HLA-DR monoclonal antibody L243 TSST-1 and 125~-labeled 35- and 28-kDa receptor species, respectively, after subtracting the strongly inhibited both 125~-labeled a parent molecular mass of uncross-linked 125~-labeled TSST-1 or SEA binding to human monocytes in a dose-dependent man"I-labeled SEA (indicated by open arrowheads). ner (Figs. 2A and 2B). In three separate experiments, L243 was consistently more potent in blocking 125~-labeled TSST-1 binding to HLA-DR than 125~-labeled SEA to the same receptor. A monocyte-specific isotype control monoclonal antibody, OKM1, had no effect on binding of TSST-1 or SEA. Two other anti-HLA-DR monoclonal antibodies, L203 and 2.06, did not block binding of either toxin. Monocyte receptors for TSST-1 and SEA were characterized further by cross-linking studies, using the bifunctional cross-linker DSS. After subtracting the molecular mass of TSST-1 and SEA (22 and 28 kDa, respectively), autoradiograms revealed two major receptor species, at

35 and 28 kDa (Fig. 3, arrowheads A and B, respectively). The specificity of binding to these two proteins was demonstrated by their absence when a 100-fold molar excess of unlabeled TSST-1 or unlabeled SEA was included (Fig. 3, lanes 2 and 6). Furthermore, no evidence of 125~-labeled TSST-1 or lZ51-labeledSEA cross-linking was observed in the absence of cells. To determine whether the receptor species were related to HLA-DR, monocytes were preincubated with L243 for 1 h before cross-linking with 1 2 5 ~ -

94 1

Can. J. Microbiol. Downloaded from www.nrcresearchpress.com by NEW YORK UNIVERSITY on 11/12/14 For personal use only.

SEE ET AL.

labeled TSST-1 or 12'1-labeled SEA. Results showed that L243 inhibited the affinity labeling of the 35- and 28-kDa bands by both 12'1-labeled TSST-1 and 12'1-labeled SEA (Fig. 3, lanes 3 and 7). Inhibition of the receptor species was not evident when monocytes were pretreated with the isotype control monoclonal antibody OKM 1 (Fig. 3, lanes 4 and 8). These results provide suggestive evidence that the 35- and 28-kDa bands represent the a! and (3 chains of the human class I1 HLA-DR antigen (Giles and Capra 1985). The role of HLA-DR antigens in the biologic function of TSST-1 or SEA was investigated by the standard mitogenicity assay. L243 or OKMl antibodies were preincubated with PBMCs before the addition of the toxin. Figure 4 shows that L243 blocked the mitogenic response of PBMCs to either TSST-1 (Fig. 4A) or SEA (Fig. 4B) in a concentrationdependent manner. No inhibition of the proliferative response was observed with OKM1. Furthermore, L243 did not affect the viability of the cells, as judged by trypan blue exclusion. No inhibition of mitogenic activity was observed with the other two anti-HLA-DR monoclonal antibodies (L203,2.06), which also did not block TSST-1 or SEA binding to monocytes (data not shown). Discussion TSST-1 and the staphylococcal enterotoxins have been called superantigens because of their ability to stimulate a broader range of T cells compared with conventional antigens. The explanation for the activation of a larger number of T cells lies in the finding that superantigens, in association with MHC proteins interact with only one chain, the (3 chain of the T cell receptor (TCR), rather-than with both the TCR a! and (3 chains, as required by conventional antigens (Marrack and Kappler 1990; White et al. 1989). The staphylococcal toxins not only stimulate T cells exclusively via the Va region of the TCR but also demonstrate preferential stimulation of T cells bearing certain TCR Va. For instance, TSST-1 has been reported to stimulate human T cells bearing primarily VD2 sequences (Choi et al. 1989; Marrack and Kappler 1990). Furthermore, the superantigens also differ from conventional antigens in that antigen processing is not required for the former and that there is no MHC allotype restriction (Fischer et al. 1989; Fleischer et al. 1989). Recently, TSST-1 and the staphylococcal enterotoxins A and B (SEA and SEB) have been shown to bind directly to the class I1 MHC antigen HLA-DR. While others have characterized TSST-1 and SEA binding to HLA-DR on transfected L cells or B lymphoma cell lines (Fischer et al. 1989; Fraser 1989; Scholl et a!. 19896), the present study is a direct demonstration of TSST-1 and SEA binding to normal human monocytes. The demonstration of specific receptors for TSST-1 and SEA on normal human monocytes has several important implications. Human monocytes have been identified as the key producers of the cytokines IL-1 and TNF in response to TSST-1 and other staphylococcal exoproteins (Ikejima et al. 1984; Jupin et al. 1988; Parsonnet et al. 1986). Both mediators may be involved in some of the characteristic features observed in TSS, such as fever and shock. Presumably, binding to class I1 MHC molecules on human monocytes may repreent the initial step by which TSST-1 or SEA stimulates the release of these monokines. Evidence to support the role of class I1 molecules in

A l5

3

i8

a

5

r

A

0

o

0.01

0.1

1.0

10

MAb (P91mL)

B --

25

Y-

20

3

i3 5

;,,

5 A FIG. 4. Inhibition of TSST-1 or SEA-induced proliferation of human PBMCs by L243. PBMCs were preincubated with the indicated concentrations of L243 or OKMl for 1 h at 37OC. Cells were then stimulated with 0.1 pg/mL (A) TSST-1 or (B) SEA and incorporation as described in Matemeasured for [3~]thymidine rials and methods. Each value represents the mean + the standard deviation for quadruplicate determinations. Background stimulation has been subtracted.

monokine secretion has been provided by Grossman et al. (1990), who reported that antibodies to class I1 molecules partially neutralized SEA-induced TNF release from human monocytes. The present study also shows that both TSST-1 and SEA bind to a common receptor on human monocytes. That the receptor on human blood monocytes represents the class I1 HLA-DR anti en is supported by the following: (i) binding of either B251-labeledTSST-1 or 12'1-labeled SEA to monocytes was blocked by preincubating cells with an antiHLA-DR monoclonal antibody, L243, and (ii) cross-linking studies revealed two receptor species with molecular masses consistent with those of the a! and (3 subunits of HLA-DR. While receptors for TSST-1 have been reported for human B lymphocytes (Mourad et al. 1989; Scholl et al. 1989a; Uchiyama et a/. 1989), the presence of specific toxin receptors on human monocytes and T lymphocytes has been more controversial. Poindexter and Schlievert (1987) demonstrated the presence of TSST-1 receptors on human T lymphocytes but did not observe specific binding of TSST-1 to

CAN. J. MICROBIOL. VOL. 38, 1992

942

human monocytes. In contrast, Scholl et al. (1989a) reported TSST-1 receptors on human monocytes, in agreement with our findings. We have recently shown that highly purified (>98% purity) human T lymphocytes did not bind iodinated TSST-1 or SEA unless in the presence of human monocyte membrane fragments (Chang et al. 1992). This finding is consistent with those of Uchiyama et al. (1989) for TSST-1 and Carlsson et al. (1988) for SEA. However, Fleischer et al. (1989) have shown that SEA and SEB can directly interact with resting human T cells or Jurkat T cell lines, since they increase cytosolic c a 2 in the absence of class I1 MHC antigen. This toxin-lymphocyte interaction presumably is receptor mediated. It is possible, therefore, that a low number of receptors or low-affinity interactions between staphylococcal toxins and T cells may not be detectable by the radioligand and binding assay, but may be detected by a more sensitive functional assay. Nevertheless, data accumulated to date suggest that monocytes and B lymphocytes are the two predominant cell types in human PBMCs that bear specific receptors to TSST-1, SEA, and SEB. Specific receptors for toxin binding on resting human T cells are apparently present but can be demonstrated only when toxins are complexed with class I1 MHC antigens. The presence and nature of these toxin receptors in T lymphocytes which function in the absence of class I1 MHC antigens remain controversial. Although TSST-1 and SEA bind to the HLA-DR class I1 molecule, there is evidence that suggests the binding epitopes may be overlapping rather than .the same. We have previously examined .the nature of common receptors for TSST-1 and SEA by cross-competition binding studies using human PBMCs from the same donors (See et al. 1990). Although SEA does block binding of radiolabeled TSST-1 and vice versa, unlabeled homologous toxin was a far better competitor than the use of the heterologous toxin. This strongly suggests that the TSST-1 and SEA may be binding to overlapping epitopes rather than to the same epitope on HLADR. Similar conclusions may be drawn from the current study, since the HLA-DR monoclonal antibody L243 was consistently more potent in blocking radiolabeled TSST-1 than SEA binding to human monocytes from the same donors (Fig. 2). Recently, Pontzer et al. (1991) have extended our findings by characterizing two binding sites for TSST-1 and SEA on the human Raji lymphoma cell line, one involving the residues 1-45 region on SEA and the 65-85 region of the MHC P chain, the other involving a different region on the SEA molecule and a separate site on HLA-DR to which TSST-1 also binds. This latter binding site could be the same overlapping epitope between TSST-1 and SEA described in the present study. Another possibility for the differences in binding between TSST-1 and SEA is that these two toxins may bind to other class I1 molecules, such as HLA-DQ or HLA-DP, in addition to HLA-DR. Scholl et al. (1989b) found that TSST-1 and SEB bind to distinct sites on HLA-DQ as well as on HLA-DR, but not on HLA-DP. Fraser (1989) and Fischer et al. (1989) noted that SEA binds to HLA-DR but not HLA-DP or HLA-DQ. Ligand analysis of our binding data revealed only one class of receptors for TSST-1 and SEA, thereby implying that HLA-DR is the main receptor in normal human monocytes. However, since the expression of HLA-DQ and HLA-DP is much lower than that of HLA-DR, one cannot totally exclude the possibility that TSST-1 and SEA may also bind to either or

Can. J. Microbiol. Downloaded from www.nrcresearchpress.com by NEW YORK UNIVERSITY on 11/12/14 For personal use only.

+

both of these molecules at low levels or under optimal conditions. The hypothesis that TSST-1 and staphylococcal enterotoxins bind to unique but overlapping epitopes on HLA-DR is intriguing. Scholl et al. (1989b), using HLA-DR and HLA-DQ transfected cell lines, have shown by crosscompetition binding experiments that TSST-1 binds to a site distinct from that of SEB. We have also demonstrated previously that excess unlabeled SEB fails to inhibit 12'1labeled TSST-1 binding to human PBMCs (See et al. 1990). These results suggest that TSST-1 and SEB bind to at least two distinct epitopes on HLA-DR, one for TSST-1 and one for SEB. In contrast, SEA has been shown to cross compete with both TSST-1 (See et al. 1990) and SEB (Fraser 1989). We postulate, therefore, that SEA occupies a binding site within HLA-DR that overlaps both TSST-1 and SEB, while the binding sites for TSST-1 and SEB appear either topographically or sterically distinct. The class I1 HLA-DR molecule has a prominent role in antigen presentation by monocytes or macrophages to T cells (Shackelford et al. 1982). Binding of TSST-1 or SEA to HLA-DR appears to be an important step in the activation of T cell proliferation. In our study, inhibition of the T cell proliferation induced by TSST-1 or SEA was observed with a HLA-DR monoclonal antibody, L243, but not with a monocyte-specific control antibody of the same isotype, OKMI. Similar reports have also documented the importance of MHC class I1 molecules in toxin-induced activation of T lymphocytes (Fischer et al. 1989; Fleischer and Schrezenmeier 1988). T cell stimulation by TSST-1 or staphylococcal enterotoxins is highly dependent on the presence of accessory cells such as macrophages (Carlsson et al. 1988; Fleischer et al. 1989). However, only HLA class I1 positive accessory cells can support T cell proliferation, since class I1 negative mutants neither bind enterotoxins nor affect T cell responses (Fleischer et al. 1989). It has been postulated that the T cell receptor recognizes TSST-1 or staphylococcal enterotoxins bound to conserved regions on HLA class I1 molecules, thereby resulting in T cell proliferation (Fleischer et al. 1989; Fraser 1989). Furthermore, it appears that the cr chain of HLA-DR is an essential binding site for TSST-1 (Karp et al. 1990). It is of interest in the present study that, in contrast to L243, the other two anti-HLA-DR monoclonal antibodies (L203 and 2.06) did not block TSST-1 or SEA binding to human monocytes (Fig. 2), nor did they inhibit TSST-1 induced or SEA-induced T cell proliferation in human PBMCs (data not shown). This is somewhat surprising since L203 has been shown to partially cross inhibit L234 binding to HLA-DR (Lampson and Levy 1980). It is conceivable from our results that the sites on HLA-DR at which L203 and L243 compete may be distinct from the overlapping epitope recognized by TSST-1 and SEA. Our finding that only L243, but not L203 or 2.06, inhibits TSST-1 or SEA binding to human monocytes further supports the epitope-specific nature of toxin binding to HLA-DR. In summary, we have shown that specific receptors for TSST-1 and SEA exist on normal human monocytes, and that the binding molecule consists of the monomorphic region of HLA-DR. The consequence of HLA-DR binding of either toxin appears to be a prerequisite for the functional activation of T cell proliferation. The overlapping nature of TSST-1 and SEA binding domains for HLA-DR has been further characterized by cross-linking studies, and by dif-

Can. J. Microbiol. Downloaded from www.nrcresearchpress.com by NEW YORK UNIVERSITY on 11/12/14 For personal use only.

SEE ET AL.

ferential inhibition of toxin binding by monoclonal antibodies to HLA-DR. The distinctive binding patterns of TSST-1 and SEA to human monocytes have additional implications. Specifically, we have recently shown that TSST-1, SEA, and SEB activate human monocytes by different signal transduction pathways (See et al. 1992a). Whether the differences in intracellular signals induced by these toxins are related to binding to unique epitopes and activation of specific monocyte protein kinases remains to be determined. The role of monocyte binding and T cell stimulation by TSST-1 and staphylococcal enterotoxins in the pathogenesis of TSS remains to be investigated. Future studies focused on receptor-mediated binding of TSST-1 and staphylococcal enterotoxins to human monocytes and T lymphocytes from normal donors and TSS patients may reveal the underlying anomalies that predispose particular individuals to TSS. Acknowledgments This research was supported in part by grant MT-7630 from the Medical Research Council of Canada, the British Columbia Health Care Research Foundation, and the Canadian Bacterial Diseases Network. Based on this work, Raymond See received a student research award in infectious diseases from the 1990 American Federation of Clinical Research, and a 1990 research trainee award from the Canadian Society for Clinical Investigation. We thank Donna Hogge and the nursing staff at the Cell Separator Unit, Vancouver General Hospital, for the provision of plateletpheresis packs, Annie Lam for preparation of the manuscript, and Linda Hui for artwork.

Brunette, N.W. 1981. Western blotting: electrophoretic transfer of proteins from sodium dodecyl sulfate - polyacrylamide gels to unmodified nitrocellulose and radiographic detection with antibody and radiiodinated protein A. Anal. Biochem. 112: 195-203. Carlsson, R., Fisher, H., and Sjogren, H. 1988. Binding of staphylococcal enterotoxin A to accessory cells is a requirement for its ability to activate human T cells. J . Immunol. 140: 2484-2488. Chang, A., Musser, J.M., and Chow, A.W. 1991. A single clone of S. aureus which produces both TSST-1 and SEA causes the majority of menstrual toxic shock syndrome. Clin. Res. 39: 36A. Chang, A.H., See, R.H., and Chow, A.W. 1992. Binding of TSST- 1 for receptors on human T lymphocytes and requirement of T cell - monocyte contact in the induction of resting T cell proliferation. Clin. Res. 40: 102A. Choi, Y., Kotzin, B., Herron, L., et al. 1989. Interaction of Staphylococcus aureus toxin "superantigens" with human T cells. Proc. Natl. Acad. Sci. U.S.A. 86: 8941-8945. de Boer, M., Reijneke, R., van de Griend, R.J., et al. 1981. Largescale purification and cryopreservation of human monocytes. J . Immunol. Methods, 43: 225-239. Fischer, H., Dohlsten, M., Lindwall, M., et al. 1989. Binding of staphylococcal enterotoxin A to HLA-DR on B cell lines. J. Immunol. 142: 3 151-3 157. Fleischer, B., and Schrezenmeier, H. 1988. T cell stimulation by staphylococcal enterotoxins. Clonally variable response and requirement for major histocompatibility complex class I1 molecules on accessory cells or target cells. J . Exp. Med. 167: 1697- 1707. Fleischer, B., Schrezenmeier, H., and Conradt, P. 1989. T lymphocyte activation of staphylococcal enterotoxins: role of class I1

943

molecules and T cell surface structures. Cell. Immunol. 120: 92-101. Fraser, J.D. 1989. High-affinity binding of staphylococcal enterotoxins A and B to HLA-DR. Nature (London), 339: 22 1-223. Giles, R.C., and Capra, J.D. 1985. Structure, function and genetics of human class I1 molecules. Adv. Immunol. 37: 1-7. Grossman, D., Cook, R.G., Sparrow, J.T., et al. 1990. Dissociation of the stimulatory activities of staphylococcal enterotoxins for T cells and monocytes. J . Exp. Med. 172: 1831-1841. Ikejima, T., Dinarello, C.A., Gill, D.M., and Wolff, S.M. 1984. Induction of human IL-1 product by Staphylococcus aureus associated with toxic shock syndrome. J. Clin. Invest. 73: 1312-1320. Jupin, C., Anderson, S., Damais, C., et al. 1988. Toxic shock syndrome toxin-1 as an inducer of human tumor necrosis factors and interferon. J. Exp. Med. 167: 752-761. Karp, D.R., Teletski, C.L., Scholl, P., et al. 1990. The a-1 domain of the HLA-DR molecule is essential for high-affinity binding of the toxic shock syndrome toxin-1. Nature (London), 346: 474-476. Kum, W.W.S., Wong, J., See, R.H., and Chow, A.W. 1992. Improved purification of staphylococcal toxic shock syndrome toxin-1. Submitted for publication. Laemmli, U.K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London), 227: 680-685. Lampson, L.A., and Levy, R. 1980. Two populations of Ia-like molecules on a human B cell line. J. Immunol. 125: 293-299. Madsen, M., and Johnsen, H.E. 1979. A methodological study of E-rosette formation using AET-treated red blood cells. J. Immunol. Methods, 27: 61-74. Marrack, P., and Kappler, J. 1990. The staphylococcal enterotoxins and their relatives. Science (Washington, D.C.), 248: 705-71 1. Mollick, J.A., Cook, R.G., and Rich, R.R. 1989. Class I1 MHC molecules are specific receptors for staphylococcal enterotoxin A. Science (Washington, D.C.), 24: 817-820. Mourad, W., Scholl, P. Diaz, A., et al. 1989. The staphylococcal toxic shock syndrome toxin 1 triggers B cell proliferation and differentiation via major histocompatibility complex-unrestricted cognate T/B cell interation. J. Exp. Med. 170: 201 1-2022. Munson, P.J., and Rodbard, D. 1980. Ligand: a versatile computerized approach for characterization of ligand-binding systems. Anal. Biochem. 107: 220-239. Parsonnet, J., Gillis, Z.A., and Pier, G.B. 1986. Induction of interleukin-1 by strains of Staphylococcus aureus from patients with nonmenstrual toxic shock syndrome. J . Infect. Dis. 154: 55-63. Poindexter, N. J., and Schlievert, P.M. 1985. Toxic shock syndrome toxin-induced proliferation of lymphocytes: comparison of the mitogenic response of human, murine, and rabbit lymphocytes. J . Infect. Dis. 151: 65-72. Poindexter , N. J ., and Schlievert , P. M . 1987. Binding of toxicshock-syndrome toxin-1 to human peripheral blood mononuclear cells. J . Infect. Dis. 156: 122-129. Pontzer, C.H., Russell, J.K., and Johnson, H.M. 1991. Structural basis for differential binding of staphylococcal enterotoxin A and toxic shock syndrome toxin 1 to class I1 major histocompatibility molecules. Proc. Natl. Acad. Sci. U.S.A. 88: 125-128. Rosten, P.M., Bartlett, K.H., and Chow, A.W. 1989. Purification and purity assessment of toxic shock syndrome toxin 1. Rev. Infect. Dis. ll(Supp1. 1): s110-s116. Schlievert, P.M. 1986. Staphylococcal enterotoxin B and toxic shock syndrome toxin-1 are significantly associated with nonmenstrual toxic shock syndrome. Lancet, 1: 1149- 1150. Scholl, P., Diex, A., Mourad, W., et al. 1989a. Toxic shock syndrome toxin 1 binds to major histocompatibility complex class I1 molecules. Proc. Natl. Acad. Sci. U.S.A. 86: 4210-4214. Scholl, P.R., Diez, A., and Geha, R.S. 1989b. Staphylococcal

Can. J. Microbiol. Downloaded from www.nrcresearchpress.com by NEW YORK UNIVERSITY on 11/12/14 For personal use only.

944

CAN. J . MICROBIOL. VOL. 38, 1992

enterotoxin B and toxic shock syndrome toxin-1 bind to distinct sites on HLA-DR and HLA-DQ molecules. J. Immunol. 143: 2583-2588. See, R.H., Krystal, G., and Chow, A. W. 1990. Binding competition of toxic shock syndrome toxin 1 and other staphylococcal exoproteins for receptors on human peripheral blood mononuclear cells. Infect. Immun. 58: 2392-2396. See, R.H., Krystal, G., and Chow, A.W. 1992a. TSST-1 and staphylococcal enterotoxins induce distinct patterns of protein phosphorylation in human monocytes. Clin. Res. 40: 53A. See, R.H., Kum, W.W.S., Chang, A.H., et al. 1992b. Induction of tumor necrosis factor and interleukin-1 by purified staphylococcal toxic shock syndrome toxin-1 requires the presence of both monocytes and T lymphocytes. Infect. Immun. 60: 2612-2618. Shackelford, D.A., Kaufman, J.F., Korman, A.J., and Strominger, J.L. 1982. HLA-DR antigens: structure, separation of subpopulation, gene cloning and function. Immunol. Rev. 66: 133-187.

Sorensen, P., Farber, N.M., and Krystal, G. 1986. Identification of interleukin-3 receptor using iodinatable, cleavable, photoreactive cross-linking reagent. J . Biol. Chem. 261: 9094-9097. Uchiyama, T., Imanishi, K., Saito, S., et al. 1989. Activation of human T cells by toxic shock syndrome toxin-1 : the toxin-binding structures on human lymphoid cells acting as accessory cells are HLA class I1 molecules. Eur. J. Immunol. 19: 1803-1809. White, J., Harman, A., Cullen, A.M., et al. 1989. The Vgspecific superantigen staphylococcal enterotoxin B: stimulation of mature T cells and clonal deletion in neonatal mice. Cell, 56: 27-35. Whiting, J.L., Rosten, P.M., and Chow, A.W. 1989. Demonstration by western blot (immunoblot) of seroconversions to toxic shock syndrome (TSS) toxin-1 and enterotoxins A, B, or C during infection with TSS- and non TSS-associated Staphylococcus aureus. Infect. Immunol. 57: 231-234. Yam, L.T., Li, C.Y., and Crosby, W.H. 1971. Cytochemical identification of monocytes and granulocytes. Am. J. Clin. Pathol. 55: 283-290.